Motor "windmilling"

[EddyWirbelstrom]

When a gas-turbine was being brought on line, its Aux Lub Oil Pump was tripped by the short-circuit element of a Multilin 239 relay ( set at 8.6 x FLC, and Time delay of Inst ).
It is assumed that the oil pressure switch did not turn off the Aux Lub Oil pump when the turbine came up to speed and the mechanical oil pump increased the oil pressure.
Is it possible that without a non-return valve in the Aux Lub Oil delivery pipe, an oil pressure buildup could have forced the Aux Lub Oil pump to rotate in the reverse direction – ‘windmilling’ the motor. When a motor is ‘windmilled’ does it draw more than its normal starting current ?

 

[itsmoked]

I can only answer the last point. If that motor were windmilling backwards when energized it can absolutely draw more than normal.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

In addition to Keith's comment;
If a motor is forced over-speed in the forward direction it may also draw excess current.
Using a 1750 RPM motor as an example;
As the motor speed is increased to 1800 RPM the current will drop. At 1800 RPM only the magnetizing current will be flowing.
As the speed is increased further the current will increase.
At about 1850 RPM the motor will be regenerating close to full load current.
As the speed is increased the current will increase
The motor has become an induction generator and it doesn't take much over-speed to make it a badly overloaded induction generator.
respectfully

 

[Marke]

Hey Keith,

When you say draw more than normal, are you suggesting the that the motor will draw more than Locked Rotor current when rotating in reverse? My impression is that the current will be similar to the locked rotor current, maybe slightly higher, but not a great deal in it. Am I wrong on this??

If you DOL start a motor that is rotating in reverse, you will have a longer start time and that can cause the protection to trip because the area under the current/time curve is greater.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[electricpete]

I would agree with slightly higher. The current vs speed curve gets almost (but not quite) horizontal as it crosses the zero-speed axis and will continue with similar slope into the negative speed region.

The motor will take longer to accelerate if it is initially reverse rotating, which would be more likely to result in an overload / time-delay type of trip than a short-circuit / instantaneous type of trip.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

You focused in on one possible cause... but there are of course many things to consider. Has the motor been bridge/meggered? Has it run successfuly since this event?

I'm a little confused by the startup sequence. Normally the electric oil pump is started first during startup of a machine. Why would the electric oil pump be switched on during startup after the main machine and its shaft driven lube oil pump was running.

"tripped by the short-circuit element of a Multilin 239 relay ( set at 8.6 x FLC, and Time delay of Inst ).". Not being familiar with the digital relays, I can't quite figure out if you had an instantaneuos trip or both instantaneous and time delay and whether your relay is set at a reasonable point. If 8.6 causes an instantaneous trip, I would think your setpoint may be a little low... the rms of (locked rotor current plus dc offset) could likely exceeed 8.6* FLA (do you have the machine kva code?). Did the relay tell you how many cycles before the trip? Were currents relatively balanced or not?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Is this a new installation or been around awhile?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[Marke]

Thanks Pete

The current vs speed curve gets almost (but not quite) horizontal as it crosses the zero-speed axis and will continue with similar slope into the negative speed region.

That is my understanding also.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[itsmoked]

Mark; That's my expectation too. It would be more correct to say the starting time will be excessive and with that the high current duration will be longer.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[EddyWirbelstrom]

Thanks for your helpful contributions. Some more information on the motor “windmilling” event:-
The 132kW Aux Lub Oil pump was running normally before the gas turbine was brought up to speed. I do not have motor nameplate data.
When the gas-turbine was brought up to speed the Multilin 239 relay tripped the Aux Lub Oil pump on “ Phase S/C” and recorded 1400A on two phases and 1900A on the third phase. The Multilin 239 does not record trip time, however the manual states that when “Phase S/C Delay” is set to Instantaneous, the relay trips within 45 mSec. ( in the past I have set Mulilin 269 relay “Phase S/C” and “S/C Delay” to 8 x FLC and Inst without causing nuisance trips on start ).
Megger test done since the incident indicate normal motor winding insulation resistance between phases and to earth.

Question:
Do the unequal phase currents measured on trip may indicate a d.c. offset which would occur with a sudden short circuit, not a gradual linear torque and current increase due to the motor acting as an induction generator as the motor speed increases above synchronous speed. ( as speed less than breakdown torque ).
Could the motor have stalled against a solid head of oil ?

 

[electricpete]

So, the electric motor had already been running for awhile when it tripped, right?

What's the motor voltage?

I can see the timing suggests the tripping of the electric motor had something to do with the startup of the machine. So you suggest "Could the motor have stalled against a solid head of oil ?"

Depends what type of oil pump. Centrifugal oil pump will draw less torque when it gets deadheaded and the motor would actually speed up toward no-load speed.... unless the centrifugal pump were deadheaded so long that it overheated and it's bearings froze up or something.

Posiitive displacement pump operating against something resembling a deadhead would generate increased torque that could stall the motor, but usually would lift a relief valve first.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[EddyWirbelstrom]

The motor rating is 132kW, 415V, 50Hz.
I will get more information on the Aux Lub Oil Pump type, motor LRA, oil pressure switch to turn off Aux Lub Oil Pump when the turbine oil pump takes over.

 

[waross]

It would also be helpful to know the normal running oil pressure developed by the Aux. pump and the normal running oil pressure developed by the turbine oil pump.
If the turbine oil pump operates at a higher pressure than the aux. oil pump then the aux oil pump may be overloaded or stalled.
It will take an examination of the lubrication piping drawings to determine whether the aux pump is liable to stall or overspeed.
respectfully

 

[electricpete]

OK. 132KW I assume is the output power (similar to how we rate motors in output horsepower). Assuming 90% efficiency and 90% power factor, the FLA would be 224A.

Then the 1400A is around 6.2xFLC and the 1900 is around 8.5*FLC. I guess that flanges up with your statement that the trip was reached at a setpoint at 8.6*FLC although it's still not clear to me whether you're talking time delay or instantaneous.

To me, 8.6*FLC sounds higher than LRA which is typically in the range 4-7*LRC... (unless it is an unusual motor... could confirm if we knew the kva code class). If so, I could see how you might get there during starting since again the LRC combines with the dc offset. But I can't see how you can get there during stall since there is no dc component. If this really is above LRC, then I would think it suggests something more than a stall due to overload. Then the possibilities might include:
- motor stopped and immediately restarted - can produce some unusual currents
- motor faulted, perhaps result of being stalled prior to that (doesn't seem likely if you have meggered and bridged, but you never know).

Also, you are right, the big current unbalance suggests something other than simple overload due to locked rotor or stalled condition.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Correction:
"To me, 8.6*FLC sounds higher than LRA which is typically in the range 4-7*FLC..."

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[EddyWirbelstrom]

I am still waiting for data from site. My assumptions are as follows :-

Aux Lub Oil OPump Motor Shaft output power = 132kW, Un = 415V, FLC = 210A, LRA = 1550A ( 7.38 x FLC ), 4 pole, 50Hz
CT ratio = 300 / 1
Multilin 239 “Phase S/C Trip” current setting = 6 x CT = 1800A = 8.57 x FLC
Multilin 239 “Phase S/C Delay” setting = Inst ( inherent relay delay when set to Inst = 20 to 45 mSec )

Oil pressure data is still outstanding.